We may joke about welcoming our robot overlords, but computers are finding ways to steal the glory from humans, especially when it comes to games.
We may have invented chess, Go, “Jeopardy,” and poker, but artificial intelligence systems have gotten so smart that our human minds are no longer the ultimate champions. Here are three prime examples of just how far Artificial Intelligence has come.
1. An AI poker ace
One of the most recent cases of computers proving their gaming superiority comes from Carnegie Mellon University, which announced that its Liberatus artificial intelligence beat a group of top poker professionals at the Rivers Casino in Pittsburgh in January.
The 20-day competition even had a fun name: “Brains Vs. Artificial Intelligence: Upping the Ante.” Liberatus didn’t just win, it wiped out the other players. The competition included 120,000 hands of Texas Hold’em. The money tally at the end found Liberatus leading the humans by $1,766,250 in chips.
Liberatus didn’t just get by on a good poker face and luck. The computer scientists in charge of Liberatus monitored the system as it analyzed its own performance every night during the tournament and then worked to fix any problems in its game. This ability to learn and adapt was key to the AI’s winning strategy.
“The best AI’s ability to do strategic reasoning with imperfect information has now surpassed that of the best humans,” said Carnegie Mellon computer science professor Tuomas Sandholm.
Carnegie Mellon noted that this kind of AI could have implications for other applications, like using a computer or smartphone to negotiate big-ticket purchases, or for business negotiations or military strategy.
2. Humans in “Jeopardy”
IBM’s Watson system may be one of the most famous computers of all time, right up there with fictional greats like HAL 9000 and Deep Thought. The year was 2011. Famous “Jeopardy” champion Ken Jennings, winner of 74 games in a row, agreed to take on Watson along with fellow champ Brad Rutter. What happened was a riveting battle between human minds and an artificial mind on one of America’s most popular game shows.
Watson had to take the same “Jeopardy” contestant test that all players are required to pass. The computer also engaged in a series of practice games against people to prepare for its starring role on TV. “Jeopardy” was a natural fit for Watson’s skill since the computer system specializes in answering natural-language questions. That means it also has to understand human speaking foibles, like puns, double meanings and slang words.
The competition involved two “Jeopardy” matches. Watson analyzed keywords in each quiz clue, identified likely answers, and then decided how confident it was in providing the correct answer before signaling its buzzer.
Despite some missed answers, Watson handily bested the human contestants in each match. IBM’s smart system “walked” away with the $1 million grand prize, which the computer company donated to charity.
Note: Watson has come a long way since he first stepped into the spotlight. Listen to this podcast about how this brilliant robot has now become an expert on cybersecurity.
IBM's Watson is now a Cybersecurity Expert
3. A chess match for the ages
Watson wasn’t IBM’s first machine-versus-human triumph. Let’s go back to the late 1990s when world chess champion Garry Kasparov was at the top of his game and agreed to take on IBM’s Deep Blue in a series of tense matches. The first set of games happened in 1996 and Kasparov, a Russian grandmaster, defeated the machine over the course of six games by winning three, drawing on two, and losing just one. IBM didn’t give up. It revamped Deep Blue and set another series for 1997.
Deep Blue came armed with more computing power for the new six-game set. Kasparov won the first game, but Deep Blue came back to claim the second. The next three games all ended in draws, which set the scene for a final dramatic showdown. The last game, which was broadcast on television, ended badly for Kasparov as Deep Blue won in 19 moves.
Kasparov asked for another rematch, but IBM chose to retire Deep Blue after its win. Deep Blue’s chess prowess was a preview of computer systems to come. At the time, it was seen as a monumental achievement in computing power. The machine could evaluate millions of possible chess positions in a mere moment. Today, chess players can test their skills against powerful software that runs on regular desktops, laptops and portable devices, but Deep Blue led the way all those years ago.
Bonus: Google conquers Go
Go is an ancient game that traces its origins back to over 2,500 years ago in China. It involves black and white stones on a board. Two players take turns moving those stones to capture each other’s pieces and claim territory on the board. On the surface, it seems simple, but it’s incredibly hard to master.
Computer minds spent many futile years trying to top human players at Go. Google’s DeepMind AlphaGo program finally conquered that mountain in 2015 when it beat European champion Fan Hui in a formal match. “This is the first time ever that a computer program has defeated a professional Go player, a feat previously believed to be at least a decade away,” Google said. AlphaGo followed that impressive victory up in 2016 by beating Lee Sedol, who is considered one of the game’s best players. It’s just one more example of how artificial minds are surpassing human brains in the gaming world.